"Kenneth" <kdw### [at] gmailcom> wrote:
>
> So, using the two *original* 1996 Daniel Mecklenburg textures from the
> "granites_original.inc" file that Thomas included in his zip files, I
> decided to run  them in v3.7 ...

I forgot to mention that the *only* minor change I made to those original
textures was to scale them smaller-- to look more like Thomas's granite.

Post a reply to this message

Maybe we should implement the Cornell Box as a standard way of posting images
for comparison - to remove all of the other variables that affect the look of an
object and texture.

http://news.povray.org/povray.unofficial.patches/thread/%3C39380E82.5F30%40wanadoo.fr%3E/

(Isn't there an include file for this?)

Post a reply to this message

"Bald Eagle" <cre### [at] netscapenet> wrote:
> Maybe we should implement the Cornell Box as a standard way of posting images
> for comparison - to remove all of the other variables that affect the look of an
> object and texture.
>
> (Isn't there an include file for this?)

Yes, the 'Cornell box' scene is in SCENES/RADIOSITY (I had to hunt for it.)

I think its particular box environment might not be the best for showing the
true colors of a textured object placed inside it, only because it has
red/green/yellow walls (and the only lighting comes from the small ceiling
'radiosity' panel.) But your idea sounds useful-- perhaps a gray sphere for the
environment? Plus at least one white light_source for phong/specular highlights?

Post a reply to this message

Just noticed that my test scene code's two light_sources are in identical
positions-- they should be...

light_source {
  0*x
  color rgb .3
  translate <20, 40, -20>
}

light_source {
  0*x
  color rgb .7
  translate <-20, 40, -20>
}

My apologies for the error. I don't think it affects the results of the overall
experiment though. (I was curious as to why my renders showed only ONE highlight
on the objects.)

Post a reply to this message

Op 15/04/2021 om 22:03 schreef Kenneth:
[snip]>
> I make no judgement as to which of the versions of 'North American Pink granite'
> has the correct 'visual' look-- I'll leave that to the granite experts ;-)
> 

It the case of the granites I sincerely wonder if the gamma issue is an 
issue at all (not to be discarded like that of course but...).

Attached is a Real World photograph of North American Pink. No scale is 
provided but, in general, the largest grains in granites do not exceed 
10-15mm [Bates & Jackson (1987): Glossary of Geology, 3rd Ed.].

Imo, this granite hue closely resembles the assume_gamma 1.0 render and 
much less the assumed_gamma 2.2 one. However, there are innumerable 
variations in hue and sizes, so who can say he is in the right and who 
in the wrong? It becomes almost trivial. Except for the /scale/ of the 
texture, where I strongly feel that a 'Real World correct' render should 
be provided to the users.

I have come to the conclusion that granites21.inc is /based on/ 
granites.inc by Daniel Mecklenberg and not an exact reproduction of his 
code. If we want those, we need to render the original file separately, 
  with the initial conditions like Ive and you have done.

-- 
Thomas

Post a reply to this message

Attachments:
Download 'north american pink.jpg' (49 KB)

Preview of image 'north american pink.jpg'

Op 16/04/2021 om 08:24 schreef Kenneth:
> "Bald Eagle" <cre### [at] netscapenet> wrote:
>> Maybe we should implement the Cornell Box as a standard way of posting images
>> for comparison - to remove all of the other variables that affect the look of an
>> object and texture.
>>
>> (Isn't there an include file for this?)
> 
> Yes, the 'Cornell box' scene is in SCENES/RADIOSITY (I had to hunt for it.)
> 
> I think its particular box environment might not be the best for showing the
> true colors of a textured object placed inside it, only because it has
> red/green/yellow walls (and the only lighting comes from the small ceiling
> 'radiosity' panel.) But your idea sounds useful-- perhaps a gray sphere for the
> environment? Plus at least one white light_source for phong/specular highlights?
> 

I agree with you. My 'environment' could serve but needs adjustments in 
that case.


-- 
Thomas

Post a reply to this message

Thomas de Groot <tho### [at] degrootorg> wrote:
>
> I have come to the conclusion that granites21.inc is /based on/
> granites.inc by Daniel Mecklenberg and not an exact reproduction of his
> code. If we want those, we need to render the original file separately,
>   with the initial conditions like Ive and you have done.
>
Yes, I agree. And in my personal opinion, the gamma 1.0 'look' is most likely
what was intended, more or less-- based on my trust of your own knowledge of
granites, and also the original comments by Daniel M, who seemed to know about
the subject himself. It would surprise me if he did run his textures with a
2.2-gamma and accepted the result-- because I would assume that "North American
Pink granite" is a well-known and 'standard' type of rock, agreed on by
geologists.  Not like 'political viewpoints', ha.

Post a reply to this message

Op 16-4-2021 om 09:36 schreef Kenneth:
> Thomas de Groot <tho### [at] degrootorg> wrote:
>>
>> I have come to the conclusion that granites21.inc is /based on/
>> granites.inc by Daniel Mecklenberg and not an exact reproduction of his
>> code. If we want those, we need to render the original file separately,
>>    with the initial conditions like Ive and you have done.
>>
> Yes, I agree. And in my personal opinion, the gamma 1.0 'look' is most likely
> what was intended, more or less-- based on my trust of your own knowledge of
> granites, and also the original comments by Daniel M, who seemed to know about
> the subject himself. It would surprise me if he did run his textures with a
> 2.2-gamma and accepted the result-- because I would assume that "North American
> Pink granite" is a well-known and 'standard' type of rock, agreed on by
> geologists.  Not like 'political viewpoints', ha.
> 
> 
Lol! My thoughts entirely. I assume that Daniel Mecklenburg intended his 
textures to be used with an added scale closely related to the 
/dimensions/ of the supporting object where a 10cm wide cube in one 
scene would need a different scale than a 10m wide stone wall in another 
scene. Caveat: if used in the same scene, both objects would of course 
need the same scale! I would argue that, by itself, any POV-Ray texture 
is "dimensionless" and needs a known support in order to be scaled 
accordingly.

I assume that the name "North American Pink" is the commercial name. As 
a geologist, I would call it a 'granodiorite' probably. The same goes 
for the other granites textures which I have been able to trace in the 
Real World.

In any case, as 'simple' users, lets be and remain creative! Also in the 
rgb/srgb matter (ouch! I did it again!). We have a saying here: "buiten 
de lijntjes kleuren" which translates as "colouring outside the lines 
(or the boundaries)" and meaning "breaking the rules" in particular 
concerning the creative process, but also applied to scientific 
research. It promotes serendipity without any doubt imo.

-- 
Thomas

Post a reply to this message

Am 4/16/2021 um 10:21 schrieb Thomas de Groot:
> Op 16-4-2021 om 09:36 schreef Kenneth:
>> Thomas de Groot <tho### [at] degrootorg> wrote:
>>>
>>> I have come to the conclusion that granites21.inc is /based on/
>>> granites.inc by Daniel Mecklenberg and not an exact reproduction of his
>>> code. If we want those, we need to render the original file separately,

>>>
>> Yes, I agree. And in my personal opinion, the gamma 1.0 'look' is most 
>> likely
>> what was intended, more or less-- based on my trust of your own 
>> knowledge of
>> granites, and also the original comments by Daniel M, who seemed to 
>> know about
>> the subject himself. It would surprise me if he did run his textures 
>> with a
>> 2.2-gamma and accepted the result

I'm can assure you that in 1996 absolutely nobody did use assumed_gamma 
1.0 for reasons that are actually quite obvious even if you do not 
understand them.
On the other hand my guess would be them someone with the pseudo "Code 
Warrior" (and its implied reference to Mel Gibson) did not use a toy 
operating system like Windows 95 which means he did use assumed_gamma 
1.8 as a consequence.


>> because I would assume that "North 
>> American
>> Pink granite" is a well-known and 'standard' type of rock, agreed on by

>>
>>
> Lol! My thoughts entirely. I assume that Daniel Mecklenburg intended his 
> textures to be used with an added scale closely related to the 
> /dimensions/ of the supporting object where a 10cm wide cube in one 
> scene would need a different scale than a 10m wide stone wall in another 
> scene. Caveat: if used in the same scene, both objects would of course 
> need the same scale! I would argue that, by itself, any POV-Ray texture 
> is "dimensionless" and needs a known support in order to be scaled 
> accordingly.
> 
> I assume that the name "North American Pink" is the commercial name. As 
> a geologist, I would call it a 'granodiorite' probably. The same goes 
> for the other granites textures which I have been able to trace in the 
> Real World.
> 
> In any case, as 'simple' users, lets be and remain creative! Also in the 
> rgb/srgb matter (ouch! I did it again!). We have a saying here: "buiten 
> de lijntjes kleuren" which translates as "colouring outside the lines 
> (or the boundaries)" and meaning "breaking the rules" in particular 
> concerning the creative process, but also applied to scientific 
> research. It promotes serendipity without any doubt imo.
> 

Well, my understanding was that this is not just about this particular 
granite file, but about a correct way to convert in general old textures 
from times were *never* assumed_gamma = 1.0 was used.

Obviously this is not the case, sorry for this misunderstanding on my 
part, but frankly you should have said so in the first place and should 
not have added a bunch of buzz words borrowed from color science into 
your header where these words make no sense at all.

So I did completely waste my time here and it seems you did not even 
bother to look at my code - as all I hear is just about comparing 
images. So here is a final image from me also produced with my code but 
with some more educated "guesses" to tweak one user parameter - based on 
the image you did show.

My conclusion: my first thought was right, so please do not mention me 
in any of your files, seriously I mean it.

No need to bother with a reply I'm outa here

so long - Ive

Post a reply to this message

oops, the image...

Post a reply to this message

Attachments:
Download 'nap.png' (496 KB)

Preview of image 'nap.png'